It is often assumed that the multiple roles of women who are mothers, wives, and employees, in addition to being primary caregivers to older relatives, compete for their time and energy, making these women vulnerable. The proposed research examines the nature and impact of these multiple roles in addition to the role of caregiver. The overall objectives are to examine (1) how well middle generation women (those aged 35 to 55) who simultaneously occupy the roles of primary caregiver to a parent or parent-in-law, mother to at lease one child residing at home, wife, and employee adapt to the demands of these multiple roles, and (2) how these roles separately, and in combination, influence women's psychological, social, and physical well-being over time. The specific aims of this research are: 1) To examine whether stress associated with each of these roles independently affects women's well-being, whether support within each role modifies role stress, and whether these role- specific processes are mediated by a sense of mastery associated with a given role; 2) To examine whether stress and support processes occurring in the mother, wife and employee roles "spillover" (either positively or negatively) to affect the caregiver role; 3) To examine whether well- being changes over time as a function of both role-specific and multiple role stress and support, and whether these changes are mediated by role mastery; 4) To examine the short-term and long-term antecedents and consequences of changes in role constellations. Role support (both positive and negative support received within a role) is hypothesized to moderate the effects of role stress (the relative amounts of problems and rewards within a role) on well-being. Role mastery (the extent to which events within a role are perceived as being within one's control) is hypothesized as a mediator in the stress, support and well-being relationship. Role losses are hypothesized to be a function of prior role stress, support, mastery, and well-being. The consequences of role changes (both losses and gains) are examined in terms of the voluntary nature of the change, the type of role or roles that change, and the prior quality of the role or roles that change. The proposed study uses a three-wave, two-year panel survey design. A purposive sample of 280 women will be interviewed in-person three times at twelve-month intervals. Ordinary-least squares multiple regression path analysis will be the primary statistical technique used.